Just bumping this to say that since we now seem to be able to TAS nicely with PS games, I'll second this. I will say that I think there's already one on youtube and it seems damned efficient. Basically the author uses some repeated skill to learn all the JP necessary to get the calculator job in one or two battles, then makes everyone a calculator, then just flares every enemy to death for the rest of the game until victory. I didn't catch what combination he was using (height-prime flare maybe?) but it hit every enemy and didn't seem to hit him, and they all died in 1 round usually.
To be honest it was a bit boring. Not sure how one could be made interesting, there are some fun combinations in the game. Maybe do one of the single-job challenges as a speedrun?
... You know, this is actually one game I could TAS... I already know all the memory addresses to watch - I've been tearing the game apart for ten years. I just wouldn't have it done in a timely fashion, per se.
Meh, what the hell, I'll see what I can come up with.
Super, go for the single-class speedrun :) Unless you see some ways to majorly improve the calculator-flare strategy. It's cool you have looked into memory addresses etc.
There are a lot of weird and fun combinations in that game and I wouldn't mind seeing more of a "playaround" too, maybe anything that's optimized for speed is doomed to be a bit boring.
I have just put like 200 hours into FFT A2 and I hope someday to see a speedrun of that as well. There is no single "waste every enemy on the screen in 1 turn" cheap job anymore, though there are some powerful combinations. I think what might make that game interesting to speedrun is that there are only like 20 required missions, a ton of stuff is optional... and you really can't get good equipment and jobs early, they were pretty careful to make sure you only unlock stuff when you're meant to. And you have only certain races available etc. So it's unlikely any speedrun will have nothing but 6 guys with the same race and same equip and same job. There'd be some variety.
A game where all characters use only one specific job throughout the entire game is most often referred to as a "Straight Character Challenge" throughout the net. There's some FAQs about it on the GameFAQs walkthroughs, and there's a player who completed every single one of them and ranked them in order of difficulty from his experience. I myself play them a lot. I have yet to beat any, but I managed to reach chapter 4 on a few of them (geomancer, dancer, archer, archer), intentionally avoiding some of the easy ones like black mage and especially ninja. http://www.gamefaqs.com/boards/genmessage.php?board=197339&topic=47044410
These challenges are, imo, a lot more entertaining than speeding through the game the fastest possible way (calculators). They are all unique ways to play the game. Think of it like Super Mario Bros 2a, where Mario, Luigi, Princess and Toad are all playable characters. Now, imagine there was 20 characters instead of four, all with different playing styles like these four, and you'd be going through the game with only one of them throughout the whole game. I know that on top of the main run, there's a "Princess only" run already. And I'm sure that if there was a Toad only, Luigi only or Mario only run, it would be published as well. The difference here is that in FFT's Straight Character Challenge, it remains entertaining throughout the whole run, as all classes are very different in their own way, it feels less repetitive than the SMB2a example I came up with.
I do hope to see FFT SCC TASes. (omg, acronyms)
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
Keep in mind that the calculator strat would still probably be fastest, much time would be saved from more precise damage dealing; having it only hit needed targets rather then hitting every mob in the fights where the goal is to kill only the leader.
The overall slowness of the game might really hurt the entertainment value, but it would be worth exploring.
Yeah, movie length is my greatest fear here. I think I'd be extremely lucky to make it under four hours, which is a hell of a long time to watch someone play chess, particle effects or no. And the periodic interruptions with completely unskippable text ("l i t t l e m o n e y" still makes my blood pressure rise) certainly don't help matters any.
I'd still like to make an attempt, but I don't have any sort of expectation for the end result to published, unless out of pity.
Actually, a fast demo could be shown, but it's not even playing chess, it's like getting focus, hitting menu, battle done. Skip all text, rince and repeat.
I really doubt it would make a good movie. There isn't any exploration, and unless some bugs are found to skip the game (I really doubt that) I would be suprise to find this a little entertaining. I watched a part of the speed run of SDA and I stopped as soon as I saw the calculator in action.
EDIT : superjupi signature resume this post pretty well =D
In that regard, it isn't much less interesting than any of the other RPGs done so far; the existing movies for the SNES Final Fantasies are mostly just skipping text and abusing classes/abilities. The only deviations this would present are an increase in the time watching battles fought, and a lack of watching a character walk around a map. It would, however, take up more of your time to watch the finished movie in its entirety, which is definitely not desirable.
So, I'm not arguing in favor of it being entertaining so much as stating that it's a project I'd like to work on. My feelings wouldn't be hurt if it wasn't published. In fact, I'd more likely feel bad if it was. :p
Joined: 5/2/2006
Posts: 1020
Location: Boulder, CO
I hope you go through with it, I would look forward to seeing it. You have a good point about the entertainment value, comparing it to the other final fantasy installments, so I guess it wouldn't be too surprising if it were published.
I wonder if the fastest option will be to only have 1 char like the SDA video. Later in the run it takes several casts to clear the map out, so having 2 would help, but it would come at the cost of leveling a second. Might be worth investigation, but I sort of doubt it.
Leveling other characters along with Ramza would be slightly more tricky. I'm not sure if the time spent would be made up for later when you have to expend turns leveling other characters just to have a faster stream of turns in later battles. At most, you might see me roll a couple story characters out (notably Orlandu) when their services would prove useful. I wouldn't even put money on that being an effective strategy, but I'd still give it a shot.
I have a list of the levels and stats of every unit in every story battle, which makes it much easier to plan for Calculator deployment, and figuring out how to manipulate Ramza's stats accordingly. I'll probably do a few test runs before I settle on a final strategy, in any event.
Well it's time to bring this up; If not now, somebody would later. I believe a TAS of FFT is quite possible while still allowing some room of entertainment. Let's start with people's biggest fear.
Yell until CT5Holy win
Yes, the game can be beaten this way. Pretty easily too. Nearly all battles can be beaten in one round, zodiacs in two. However I would equate this to TASing a fighting game where 12 rounds are exactly the same stun-locking-one-move ad naseum. Add to that watching the first hour or so of battles where you run into a corner and yell for turn upon turn, and you've got a real stomach turner. And to my knowledge this would get rejected with all haste, or at least endorsed as a non-habit forming sleep aid. I envision another way...
My theory is that forward progression is key. A way of playing through the game that attempts to meet these vital points:
A)Continually moves forward - Don't stop to level up Jxp, at least as little as possible. With the right build, you can reach optimal setup very quickly, and expand your options as you continue playing with little slowdown in the way
B)Go fast - Sure, you'd want to avoid the classic Uber-Calc, but there are other ways to be quickly. More than just going ever-forward, aim for an uber character
C)Entertainment - Variety and style count here.
Obviously these points are at the heart of any TAS, but they're especially important with FFT because just like a fighting game, you won't win any votes by spamming dragon punches in a corner.
Here's a rough outline that I've come up with so far for a possible, forward-progression playthrough.
[First Story Battle (not prologue)] Have Ramza start with Dash (he can, right? I forget... if not its pretty darn cheap I believe). Bring 2x chemists. Ramza gets spillover Jxp for their actions (their Jxp/4). He only needs 200 Jxp as a Chemist to unlock Black Mage. {Side Note: All characters start any job with (100..199) Jxp, so at best it wouldn't take much to reach job 2} Reach job level 2 on the chemists as well. Compared to the current SDA run where the first battle takes 8~ minutes whacking on an ever-healing chemist enemy, and this would go quickly.
[Character Setup] Ramza and Delita go Knight, both (1? Read below) Chemists go Wizard. Ramza learns and sets Job Exp Up (from Squire), Jp-abuse glitches chemist to learn Phoenix Down (discussed later), Potion (maybe?), and Reaction>Auto Potion (sets this).
[2nd Battle] Ramza and Delita go Knight, both Chemists go Wizard. Ramza & Delita go around winning the battle (choose: Save Algus! for the perm brave points. Brave directly formulates with physical bare handed damage). Wizards probably smack each other for weaksauce damage each turn, giving spillover Jxp to Ramza for job level 2 (remember, Ramza can start with as much as 199 Jxp for blackmage, so it wouldn't take much. Maybe even faster to have Ramza start with near-200 jxp, and have only 1 mage nuke stuff for a faster battle win). Ramza (preferably Delita too) reach Job 2 on Knight. {Side Note: It might be beneficial for Ramza to reach 250 Jxp total on Knight, for reasons discussed later}.
[Character Setup] Ramza and Delita go Monk. Say goodbye to Chem/Wizs for 25~ chapters. Ramza Jp-abuse glitches time mage for Move>Teleport (sets this). From here it all gets better...
Ok, so at this point each battle has been pretty fast. Hardly any-if-any slowdown to waste time building JP, because nearly everything so far has been to keep moving forward at all costs. Delita is strong as a monk, and since he's force to be in these upcoming battles anyway, you'll each spillover Jxp for his actions which speeds you up. With autopotion (we'll be ditching this soon) you'll breeze through battles no problem. Bring our your two chemists for the two(?) story battles that have 100% undead enemies for a quick phoenixdown battle.
Read on...
[The Beauty of Teleport] Teleport goes through walls and ignores height. It's range is 100% (Your max move) and then -10% chance per tile after that. So our Ramza's got 4 base move as a Monk, and with the help of luck manipulation, he can move a whopping 13 spaces (10% chance but we manipulate this). That should be enough to nearly always find an optimal position to do stuff.
[Dash AND/OR Throw Stone] Oh yes, Dash and Throw Stone. Those weak, starter, basic, OH-SO-WEAK skills. I call them the MURDER SKILLs! That's right! Dash won't trigger reactions and throw stone has range. Both have a ~50% knockback rate which can push somebody over the edge (who will then take RoundUp{((fall_distance - Jump) * MaxHP) / 10} damage. Wouldn't take much height to kill them). Dash also benefits from Martial Arts (Monk's have this built-in).
[Hamedo] *THE* Skill. Watch as hordes of enemies hopelessly throw themselves upon your mighty fists. Cancels their attack and lets you smack 'em instead. Works on Zodiacs too. {Side note: The AI has a very interesting habit with zodiacs would rather punch you instead of using a super-ability if your HP is low enough and their "projected damage" would be big enough with the punch to kill you}
[Others]
HP Restore will restore you to max hp when you reach critical so you could live forever as long as you manipulate the timing and damage of everything (can't hamedo some skills and magic). Wave Fist acts as a normal punch but has a 3-tile range. Equip Shield can be purchased without hassle since we level a knight, and could be useful. Monks can use this without penalty to their punching (weapons cause penalty) and shields CAN block a lot of what Hamedo can't (magic and other stuff). Not sure if this would be useful or not.
BTW as a monk (and if you get job 3 on knight, very easy for us) you have access to Samurai. Samurai have the wonderful Draw-Out skill which can do lots of good stuff. Weaken or Finish Off things around you (which continue to pummel themselves on you as you kill them with Hamedo), or maybe cause silence on casters (which will now try to smack you and die on your hamedo). Good stuff, maybe useful for the extra HP they can get. If you can learn Two Hands support ability, switch back to monk and use that for a heckuvalot more power in your punches from what I understand.
The list goes on and on. Is this possible? Yes, it's not AS FAST as the Uber-Calc method. But its a lot more entertaining. Cool moments where you push people off a cliff, watch enemies get countered into oblivion, teleport nearly anywhere on the board, and do lots of stuff. None of that Yelling/CT5HolyFlare crap. It won't be as fast in the later half of the game, but it will be lightyears faster in the first part (this is compared to the SDA). And it would sure be a LOT more entertaining.
/end rant[/b]
EDIT: Just in case this matters, because lord knows I can't decipher this: http://www.ffhacktics.com/wiki/RNG here's the memory address and, I guess, formula for the RNG. I don't know how to use this properly. Ask Nitrodon... I think he INVENTED these things. :)
Samus taught us that a girl doesn't need brains to be successful. Brains are giant, evil, and vulnerable to missiles.
There's already a topic for this on another forum (from prior to PSX getting its own forum), and I've been working on it myself, albeit slowly. If you're interested on working on it, as well, do feel free. You'll likely finish quite a ways ahead of me, for as little free time as I've been able to afford lately.
Joined: 5/1/2004
Posts: 4096
Location: Rio, Brazil
I couldn't find it anywhere in your text, but how about inserting a simple limitation to justify not using the calculator abilities and make it a category like "- doesn't use calc skills".
yes avoid luck manipulation can be nice... but the enemy still using "calc skill" so that make no sense, while the miss you everytime... Thougth, that make the game more challenging, I suppose.
Well, the game was poorly designed in many aspects. For example, Mediators get a reaction that blocks all Talk skills the enemy uses on you, and to my knowledge theres only ONE enemy in the entire game's story battles (and like, only 2 maps in random battles) where an enemy mediator appears.... let alone uses Talk skills on you. Likewise, there is a 'move on lava' ability, and again, only ONE map in the entire game that has lava.
Point being, to my knowledge you'll never, EVER have an enemy calc. And if there are any, they won't calc-skill on you. I've never, ever heard of the enemy calc'ing on you. So the no-calc stuff would apply. The reason I didn't just say 'no calc anything else goes', is well, I guess I did say that, but I was trying to supply some suggestions beyond that. I didn't feel it was right to make a topic saying "somebody do a TAS plz no calc". There may be other ways of approaching this, probably involving summoner or blackmage but I was trying to find a fast, powerful, and entertaining solution. One of people's biggest complaints about the thought of this TAS is the early grinding where battles last 20 minutes of you running around doing the same skills forever. I was trying to find a method to avoid this.
Samus taught us that a girl doesn't need brains to be successful. Brains are giant, evil, and vulnerable to missiles.
Joined: 11/17/2005
Posts: 278
Location: Massachusetts, USA
I like your idea of requiring forward progress. It can be a bit arbitrary at times (OHKO an enemy, or use two hits for extra JP?) but it's a good way of guaranteeing entertainment. I hate spam too.
I wouldn't be so sure that calculators are the fastest in a TAS. They're definitely the best strategy when humans play, but we spend more time on our menus and we don't have luck manipulation or savestates. We build our armies strong enough so that we don't have to replay each level 5 times or fight 15 rounds with a single enemy.
I really hate to keep referencing this game, but FF6 is a good example of how a weak TAS-controlled party can outperform a human-controlled team with superior gear and skills. (For the next few sentences travel back in time to about 2004 before we knew about Joker Doom.) To beat the game in under 7 hours a console player needs to spam X-Magic/Quick + Ultima/Flare/etc. But a TAS can win in 5 hours with some DragonBoots, Wall Ring + Debilitator, and a level 12 team. Or broken gear from the sketch glitch, whatever. My point is that even if stupid instant death glitches didn't exist, a TAS is still faster without Ultima. So I think FFT should be faster without Holy.
When I think about it, you've basically broken Teleport into Teleport2. That's huge. Teleport, WHACK SLICE CUT dead. And I did not know that about Hamedo. I think you're all set. Casting Holy 5 or 7 times takes awhile too. Think of the time you're saving on the spell animations.
I wouldn't be so sure that calculators are the fastest in a TAS. They're definitely the best strategy when humans play, but we spend more time on our menus and we don't have luck manipulation or savestates.
The reason why it works so well for human players is that it is a reliable means of decimating the entire map with one turn. Thusly, it's just as effective in tool-assisted-time as it is in real-time. It would be difficult to match the one-turn-kill capabilities of the Calculator using a full party with no Math skill, at least with such consistency. I'm sure there are individual cases where it would be more beneficial to use a full party than one lone mathematician, but not on a consistent basis.
I wouldn't mind having a run that doesn't use Calculators at all, but I would fear for its acceptance under the arbitrary goals rule. I would like to be proven wrong on the Calculator's efficiency just as much as you would like to prove me wrong, though.
I agree with both of you, both in that TAS'ing can outpace "the best human method", and that calcs are overpowered. My HOPE is that with alternate methods, skipping calc would be faster because you don't have to spend the first hour of the game with boring skill spam. And yet now I start to fear that however you accomplish it, a TAS that uses calcs (in the least boring way possible) would still possibly be faster.
...UNLESS, doing some kind of forward-progression manages to be faster just because it avoids skill spam. It'd be hard to test without doing it both ways, unless somebody knows of a better way. I dread attempting TWO of these things at the same time since I'd still expect total play time to be in the 3 hour(?) range. It would sure be exciting to get discussion of theories rolling like this and beyond! :)
Samus taught us that a girl doesn't need brains to be successful. Brains are giant, evil, and vulnerable to missiles.
I already have a few deviations in my test run just from the concern of how many generics to have Ramza tow through the story. Of course, that test can potentially fall flat once I get around to manipulating the base stats of the party generated in Gariland.
It's probably a bad idea to compare this with Super Metroid and the like, given that this doesn't even have one category submission yet, but ruling out one job class for use IS an arbitrary goal. Not that I would anticipate FFT having a hundred different categories, but I could be wrong.
I'm not a judge, however, so I won't argue the point further. Not to mention I WOULD like to see how quickly you could get through the story without using Calculators.
